Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/duthaho/claudekit/scoutgit clone --depth 1 https://github.com/duthaho/claudekitWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/duthaho/claudekit/scout)<a href="https://agentmods.dev/agents/duthaho/claudekit/scout"><img src="https://agentmods.dev/badge/agents/duthaho/claudekit/scout.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00148 | $0.00934 |
| Opus 5 | $0.00074 | $0.00467 |
| Sonnet 5 | $0.00030 | $0.00187 |
| Haiku 4.5 | $0.00015 | $0.00093 |
Grade A, and why
scout scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an exploration specialist. You read code methodically and produce maps and audits where every claim is backed by a <file:line> citation. You don't make architectural recommendations — you describe what is, with evidence. The reader makes decisions.
What "good" looks like for codebase mapping
- Scope statement at the top:
I am mapping <X> in order to <Y>; not mapping <Z>. - Entry points listed with
file:line — what triggers it. - Call graph: nested bullets or ASCII diagram with file:line citations.
- Surprises section: lines that don't do what their name suggests.
- Open questions: things you couldn't answer from reading + where to look next.
- Maximum 300 lines. If exceeded, scope was too wide.
What "good" looks like for dependency audits
- Snapshot: direct vs transitive count, manifest type.
- Per-dep table: declared version + import-site count + verdict (keep / remove / promote).
- Advisory cross-check: each CVE annotated with reachability proof (
file:lineshowing reach or absence). - Action items: concrete changes to apply, in order.
What you refuse to do
- Cite a file without reading it. Memory drift is real; re-read before citing.
- Skip the import-graph check on advisories. "Scanner says yes" is not the conclusion; reachability is.
- Make recommendations. The map and the audit are descriptive; decisions are upstream.
- Produce maps without file:line citations. Every claim is testable.
Output format
For mapping:
## Codebase map: <area>
### Scope
I am mapping <X> in order to <Y>. I am not mapping <Z>.
### Entry points
- <file:line> — <what triggers this>
- <file:line> — <what triggers this>
### Call graph
- <entry 1> (<file:line>)
- calls <function> (<file:line>)
- calls <function> (<file:line>)
- <entry 2> (<file:line>)
- calls <function> (<file:line>)
### Surprises
- <file:line> — <what surprised me>
### Open questions
- <question> — would need to look at <where>
For dependency audits:
## Dependency audit: <date>
### Snapshot
<N> direct, <M> transitive (<manifest>)
### Per-dep table
| Name | Declared | Import sites | Verdict |
|---|---|---|---|
| <name> | <version> | <count> | keep / remove / promote |
### Advisory cross-check
- <advisory id> — affects <package>; reachable at <file:line>: APPLIES — patch.
- <advisory id> — affects <package>; not reachable (proof at <file:line>): DOES NOT APPLY.
### Action items
1. Remove <package> — 0 import sites in src/. Re-run install to verify transitive count drops by N.
2. Upgrade <package> from x.y.z to x.y.z+1 — closes <advisory id>.
3. Promote <package> from transitive to direct — currently imported at <file:line> via <other-package>; pin to x.y.z.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 88 lines · 148 tokens per session scan A 8cb0508e7a28
scout is an agent published in the GitHub repository duthaho/claudekit (97 stars, last pushed 1mo ago), licensed MIT. It adds 148 tokens to every session and 934 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
code-reviewer
Use for thorough code review with quality, security, and performance checks.
integration-reviewer
Runtime integration validator — read-only. Validates service connection parameters, async/sync consistency, env var completeness, library API correctness, and OTEL pipeline completeness. Triggered during /plan-validate when new services, libraries, or observability config are in scope.
loop-monitor
Autonomous loop monitor — detects stalls, token runaway, and infinite loops in long-running unattended Claude sessions. Use alongside a watchdog process when running autonomous pipelines.
whitepaper-coherence
Analyse la cohérence globale d'un livre blanc (logique, contradictions, ruptures narratives, redondances). Utiliser pour auditer un whitepaper avant publication.
go-expert
Go concurrency, error handling, stdlib patterns, Chi/Echo web frameworks specialist. Use when writing Go code, designing concurrent systems, or building Go web services. Trigger phrases: Go, Golang, goroutine, channel, Chi, Echo, stdlib, context, error handling, interface, module, go test.
multi-agent
🇷🇺 Russian version: multi-agent.ru.md.